A New Model for Prediction of Heart Failure Hospitalizations

Abstract

Results of the IMPEDANCE-HF trial have shown that lung impedance (LI)-guided treatment reduces hospitalizations for heart failure (HF) and all-cause mortality. The aim was to prove that degree of pulmonary decongestion during HF admission is a power predictor for 30 days readmissions.Study population included 266 patients, with HF and LVEF ≤ 35% in NYHA class II-IV. Patients were randomized (1:1) to a control group treated by clinical assessment and a monitored group whose therapy was guided by LI and were examined monthly in outpatient clinics. The degree of pulmonary congestion at each visit was assessed by the index ΔLIR and calculated as ΔLIR = {[currently measured LI/normal baseline (calculated for each patient)] −1} × 100%. The follow-up period was 642 years in the monitored group and 510 years in the control group (P = .001). Groups were similar with respect to baseline characteristics. There were 249 and 475 HF hospitalizations (mean 0.39 and 0.93 per × year of follow up) in the monitored and control groups, respectively (P < .001). There were 56, 22 and 73, 51 all-cause and HF-associated death in the monitored and the control groups, respectively (P < .001). Hazard Ration (HR) for HF hospitalizations was higher in control group [HR = 2.56, 95% CI: 1.8–3.7, P < .00001], (Anderson-Gil Model). The difference between ΔLIR at admission and at discharge was used to assess improvement in the pulmonary congestion. The improvement in ΔLIR was graded by quartiles (Q1 < 4.9%; Q2: 4.9–9.7%; Q3: 9.8–14.7%; Q4 > 14.7%). Improvement in NYHA class at discharge was graded as 1 or 2, in weight as <3 or >3 kg, leg edema as 1–3 and level of lung rales as 1 or 2 were also used to assess probability of HF readmissions. Rate of readmissions according to Q1-4 are presented in Table 1. HR's for HF readmissions at the different levels of decongestion are presented on Fig. 1. Improvements in NYHA, weight, lung rales and leg edema during HF hospitalization had a week predictive power for readmission. Conclusion: The extent of pulmonary decongestion during HF admission is a power predictor for readmissions within first 365 days for both groups. LI-guided treatment significantly reduced HF readmissions. Discharge of patients of both groups at the Q1and Q2 level of pulmonary decongestion resulted in an unacceptable readmission rate.Table 1Frequencies of Heart Failure readmissions according different levels of pulmonary congestion improvement (quartiles during index hospitalization)Fig. 1Cumulative Hazard Ratio for heart Failure (HF) readmissions according the degree of pulmonary decongestion (Q1-Q4).View Large Image Figure ViewerDownload Hi-res image Download (PPT)